This application is submitted in response to NIMH research topic J1, neuroscience and basic behavioral science in the SBIR solicitation and PA-04-086, development of high throughput tools for brain and behavior. We aim to alleviate the bottleneck in drug discovery for cognitive dysfunction with CognoCube(tm), an automated high-throughput system of behavioral analysis of cognitive function. CognoCube(tm) will comprise innovative automated testing, computer vision, and machine-learning algorithms to capture and integrate behavioral effects of drugs affecting cognitive function. CognoCube will capture many cognitive domains of high degree of homology between humans and rodents that are relevant in CNS disorders such as schizophrenia, dementia, attention-deficit disorder and ageing. This system will provide rapid and accurate indices of the therapeutic potential of novel cognitive enhancers. Phase I will be focused on proof of concept, namely, capture of behaviors relevant to targeted cognitive domains, sensitivity to pharmacological manipulation and robust bioinformatic analysis. The CognoCube(tm) system will expand its capacity during Phase II of the project to be able to assess a library of 1000 compounds per year. With libraries chosen to provide a high likelihood of CNS activity we expect to be able to detect up to a dozen new compounds per year with CNS activity in the area of cognition. Those compounds will be further assessed with standard tests of cognition and their mechanism of action will be studied through pharmacological means and binding profiling. Commercial value: The population that could benefit from new and improved cognitive enhancers is estimated at more than 30 million in the United States alone. Major challenges for drug discovery in this area include the lack of reliable and efficient in vivo testing systems. Although standardized objective test batteries have been developed for evaluation of human cognition, there is no such standard battery for rodents. Our automated cognitive testing system for mice will provide superior reliability, greater statistical power and higher throughput than standard cognitive tests and, with its power to screen libraries in vivo, will discover novel compounds with efficacy in cognition. [unreadable] [unreadable]